Knowledge Panel shows wrong info: how to fix sources (without hacks)

If your Knowledge Panel shows the wrong job title, a weird bio, or facts you did not choose, here is the uncomfortable truth:

Google is not “reading your site wrong”. It is trusting another source more.

That is not a moral judgement. It is a risk decision.

Step 0: stop thinking “schema will fix it”

Schema helps resolve ambiguity. But it cannot override source hierarchy.

If a directory looks like a stable “people database”, it can dominate early.

Your job is to change the system’s confidence:

• reduce contradictions
• increase agreement across independent sources
• make your canonical person page citeable

The model behind this is here:

• How to get a Knowledge Panel for a person (system model)

Step 1: identify which sources Google is using

In practice, you can infer it from the SERP:

• which domains rank for name + alias
• which snippets repeat the panel’s wording
• which profiles appear in the “profiles” row

If one domain’s snippet reads like the panel, that domain is acting as a primary biography source.

Step 2: decide what your canonical “person source” is

Pick one page and treat it as the canonical biography source on your domain.

For Casinokrisa, that page is:

• /person/mikhail-drozdov

Supporting pages that reduce ambiguity:

• /press
• /research
• /person/mikhail-drozdov

Important: do not create 5 competing “bio” URLs. That creates ambiguity.

Step 3: make the canonical page speak “Knowledge Graph language”

KG likes banal, repeated assertions:

• “Mikhail Drozdov (Casinokrisa) is …”
• “Founder of …”
• “Official website …”
• “Primary profiles …”

It looks boring. It works because it lowers risk.

Step 4: fix the dominant external source (if it’s wrong)

If a directory is dominating and contains wrong or outdated info, you have two options:

1. Fix the directory (best when possible).
2. Out-compete it with stronger agreement elsewhere (slower, but works).

If the directory allows edits, do the boring thing:

• update job title to match your current positioning
• ensure the site link points to your canonical person page (not a random homepage)
• ensure the alias spelling matches

Step 5: align Tier‑1 profiles (this is where panels are won)

Pick 1–2 sentences and copy them into:

• LinkedIn
• Crunchbase
• GitHub
• X
• YouTube
• Telegram

And make sure each profile links back to:

• https://casinokrisa.com/person/mikhail-drozdov

Panels love agreement across independent sources.

Step 6: build one “independent profile” (optional but powerful)

If you already have Medium/Substack/HackerNoon profiles, great. The key is not “more profiles”. The key is a stable biography page on a non-owned domain that links back to your canonical person page.

This creates the triangle:

Your site -> independent profile -> social platforms

That triangle is what makes the person entity “real” to systems.

How to keep the signal clean

The strongest Knowledge Panel setup is not a large public archive. It is a small, stable evidence stack.

For a person entity, keep these pages indexable:

• the canonical person page
• the evidence or press page
• formal research/work pages
• a small number of articles that prove topical expertise

Keep these as support-only unless they have strong independent demand:

• social profile lists
• video appearance archives
• tag pages
• glossary definitions
• duplicate biography or "about me" variations

This matters because Knowledge Graph systems look for agreement. If five pages on the same domain all try to be the biography source, they create a subtle contradiction. If one page owns the person facts and the other pages link back to it, the graph is easier to resolve.

For Casinokrisa, the clean hierarchy is:

• person source: /person/mikhail-drozdov
• evidence source: /press
• research source: /research
• model source: Entity-based SEO

That is the structure to keep stable while external profiles and third-party mentions catch up.

What to submit first

Do not submit every related URL at once.

Submit the smallest set that explains the entity clearly:

• the canonical person page
• the evidence page
• the research page
• one entity/Knowledge Panel explainer
• one indexing/visibility pillar

Then wait for recrawl and compare the cached snippets, SERP snippets, and URL Inspection status. If Google starts using the correct title, bio, image, or source relationship, expand slowly. If it does not, fix the strongest source first rather than adding more pages.

This keeps the feedback loop readable. If twenty pages change at once, it becomes impossible to know which signal helped or harmed the panel.

Step 7: stop changing identity signals for 6–8 weeks

This is the part most people fail.

If you keep changing:

• the handle list
• the bio line
• the canonical URLs

…you keep creating churn, which looks like risk.

Stability is the cheat code.

Quick checklist

• Canonical person page exists and is linked from nav/footer.
• Evidence page exists and links to strong external references.
• Research page exists if there are citeable work artifacts.
• Posts reference the same Person entity as author.
• External profiles link back to the canonical person URL.
• Dominant directory source is corrected (or you deliberately out-compete it).